Geomagnetic disturbances (GMDs) pose a risk to a power grid because the resulting geomagnetically induced currents (GIC) can lead to significant disruptions in, or a complete collapse of, the grid in a short period of time. Depending on the severity of a GMD and the resultant GIC effects on the grid, specific corrective actions need to be taken to prevent damage to the grid. However, predicting the severity of a GIC event in real-time to detect and assign severity to GIC events presents numerous challenges.